This Is an Author's Original Manuscript of an Article Whose Final And

This is an Author's Original Manuscript of an article whose final and definitive form, the Version of Record, has been published in the Annales de la Société Entomologique de France (Published online: 24 May 2013) [copyright Taylor & Francis], available online at: <a href="/goto?url=http://www.tandfonline.com/10.1080/00379271.2013.774741" target="_blank">http://www.tandfonline.com/10.1080/00379271.2013.774741 </a>. 12Erebia serotina Descimon & de Lesse 1953 (Lepidoptera: Nymphalidae), a recurrent hybrid between two distantly related species. 3456François Michel(1), Emese Meglécz(2), Jean-François Martin(3), Henri Descimon(2, 4) 78<ul style="display: flex;"><li style="flex:1">(1) </li><li style="flex:1">Centre de Génétique Moléculaire, CNRS, 1 Avenue de la Terrasse, 91190 Gif-sur- </li></ul>Yvette, France. E-mail : <a href="mailto:[email protected]" target="_blank">[email protected] </a>9<ul style="display: flex;"><li style="flex:1">(2) </li><li style="flex:1">IMBE UMR 7263 CNRS IRD, Case 36, Aix-Marseille University, 3 Place Victor </li></ul>10 Hugo, 13331 Marseille cedex 3, France. E-mail : [email protected] 11 12 <ul style="display: flex;"><li style="flex:1">(3) </li><li style="flex:1">European Biological Control Laboratory – USDA – Campus International de </li></ul>Baillarguet, 34980 Montferrier/Lez, France. E-mail : [email protected] 13 14 15 <ul style="display: flex;"><li style="flex:1">(4) </li><li style="flex:1">2 Bld. Rougemont, 13012 Marseille, France. E-mail: <a href="mailto:[email protected]" target="_blank">[email protected] </a></li></ul>16 17 Correspondence and reprints : Henri Descimon, 2 Bld. Rougemont, F-13012 Marseille, France. 18 19 20 21 22 23 24 25 Tel: +33 (0)491874156; Running title : Recurrent crossing of distantly related taxa 126 27 28 29 30 31 32 33 34 35 Abstract Erebia serotina was described in 1953 as a scarce, low elevation endemic Pyrenean species flying late in the season. At least 34 individuals are known from various locations. However, the absence of females suggests a hybrid origin, and E. epiphron and either E. pronoe or E. manto have been proposed as possible parents. Electrophoretic analysis of five allozyme loci and sequencing of three mitochondrial DNA segments and one nuclear gene now demonstrate that E. serotina results from the cross between E. epiphron females and E. pronoe males. We have used our and previously published sequence data to generate a molecular phylogenetic tree of the genus Erebia which shows that these two species are only distantly related. The question of why they happen to hybridize on a seemingly routine basis is thus raised. 36 37 38 39 40 41 42 43 44 45 46 Résumé Erebia serotina a été décrit en 1953 et a été d’abord considéré comme une espèce endémique des Pyrénées, tardive et de basse altitude, dont au moins 34 individus ont été récoltés jusqu’ici dans diverses localités. Cependant, l’absence de femelles suggère une origine hybride, avec E. epiphron et soit E. manto, soit E. pronoe comme parents possibles. L’analyse électrophorétique de cinq loci enzymatiques et le séquençage de trois segments de l’ADN mitochondrial et d’un segment de gène nucléaire démontrent qu’E. serotina est le résultat du croisement entre femelles d’E. epiphron et mâles d’E. pronoe. Sur la base d’un arbre phylogénétique présenté ici, ces deux espèces ne sont que lointainement apparentées au sein du genre Erebia. La fréquence notable et la régularité de leur hybridation dans les Pyrénées n’en est que plus intrigante. 47 48 Keywords Interspecific hybridization, Haldane’s rule, hybrid DNA, allozymes, Erebia phylogeny. 49 2 50 51 52 53 54 55 56 57 58 1. Introduction It is not an exceptional event that a hybrid should be described as a species. Among butterflies, noteworthy instances include Ornithoptera allotei (Rothschild 1914), now believed to be a hybrid between O. priamus urvillianus Doubleday 1847 and O. victoriae Gray 1856 (Rousseau-Decelle 1939; Straatman 1976), Papilio nandina (Rothschild & Jordan 1901), subsequently proposed to be a hybrid between P. dardanus Yeats in Brown 1776 and P. phorcas Cramer 1775 (Clarke 1980), and Lysandra cormion, which was described by Vladimir Nabokov (1941) and is most probably a hybrid between L. coridon Poda 1761 and Polyommatus daphnis Denis & Schiffermüller 1775 (Schurian 1989). 59 60 61 62 63 64 65 66 67 In other cases, hybrid individuals were attributed to a species already described and of similar aspect. A striking instance is provided by Lysandra syriaca italaglauca, which was described by Vérity (1939) from Italy as a form of a species from the Near East. This butterfly was subsequently shown by a karyological study (de Lesse 1960) to result from a cross between L. coridon and L. bellargus. Paradoxically, a hybrid of the same parentage, which had been described as a true species under the name polonus by Zeller (1845), and is not uncommonly encountered in some parts of Europe, has long been regarded as a L. coridon x L. bellargus hybrid by European lepidopterists (e.g. Jacobs 1958), a fact which was confirmed through chromosome examination by de Lesse (1960). 68 69 70 71 72 73 Erebia serotina was described by Descimon & de Lesse (1953) on the basis of two male individuals captured in September 1953 by HD around Cauterets in Central Pyrénées (tab. 1), at two sites located 2.5 km apart (so that these two individuals were probably not issued from the same brood). The following year, two additional males were collected and dissected to allow H. de Lesse to perform a karyological study (Descimon & de Lesse 1954). Actually, HD first considered it likely that these individuals were hybrids, but a thorough examination 3 74 75 76 77 78 79 80 81 82 of the “pro” and “con” arguments with de Lesse eventually led both authors to favour the hypothesis of a new species. Indeed, the morphological characters did not suggest a hybrid origin and the features of spermatogenesis seemed normal, which is usually not the case in hybrids. This impression was further confirmed (pers. comm. to HD) by Z. Lorković, whose pioneer studies constitute (jointly with those of H. de Lesse) the foundation of modern work on chromosomal speciation in Lepidoptera. E. serotina was thus considered to be a scarce, late flying species, perhaps a relative of E. christi Rätzer 1890. The latter species was the rarest among previously known European Erebia, and exists only as scanty colonies in a very restricted region of the Simplon Alps. 83 84 85 86 87 88 89 90 From 1953 to 1960, Henri and Robert Descimon captured 18 E. serotina male individuals, all of them around Cauterets and in September, except for one, collected on 23-VIII-1955. A morphological study carried out on these specimens (Descimon 1963) confirmed the previous conclusions. Four individuals were collected at Cauterets in 1968 and 1970 by LalanneCassou & Lalanne-Cassou (1972), and two additional ones in another valley of the French Pyrénées, above Arrens, by Louis-Augustin (1985). Lantero & Jordana (1981) found it also on the Spanish side. The circumstances of the catches were similar: single individual captures, at moderate elevation and late in the season. 91 92 93 94 95 96 97 However, Bourgogne (1963) had drawn attention to the absence of female individuals among the first 18 collected butterflies, recalling that such an imbalanced sex ratio is often observed in interspecific hybrid broods. He pointed out that the haploid chromosome number of E. serotina (N=18) was intermediate between those of E. epiphron Knoch 1783 (N=17) and E. pronoe Esper 1780 (N=19) and that its morphological features could also be interpreted as intermediate between these species (fig. 1). He concluded that E. serotina was likely to be a hybrid between E. epiphron and E. pronoe. Later on, Warren (1981) proposed again that E. 4 98 99 serotina was a hybrid, but between E. epiphron and E. manto Denis & Schiffermüller 1775, although the latter hypothesis was not supported by karyological data. Lalanne-Cassou & Lalanne-Cassou (1989) noted that E. serotina sometimes bears androconial scales, whereas E. epiphron and E. manto never do; by contrast, E. pronoe has well-developed androconiae. These facts supported Bourgogne’s, rather than Warren’s hypothesis. More recently, Chovet (1998) obtained from artificial pairing between E. epiphron and E. pronoe three adult butterflies that were rather similar to the wild-caught individuals. 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 Actually, neither morphological evidence nor karyology, nor ecological data, nor even breeding provide indisputable criteria to distinguish hybrids from representatives of a “good” species. The case of Ornithoptera allotei (Schmid 1970; Blandin 1973; Straatman 1976) illustrates the fact that the hybrid versus species debate can go on endlessly in the absence of solid genetic data. However, with the advent of molecular genetics and its formidable tools, such problems are no longer intractable. Protein electrophoresis and DNA technology offer insight into the very genotypes of the organisms under scrutiny. Codominant nuclear markers, diagnostic for the putative parent species, should display heterozygote patterns in a hybrid. Moreover, additional information can be expected from mitochondrial DNA (mtDNA), which is maternally inherited, so that its sequencing should make it possible to identify the maternal parent of hybrids (that is, as long as candidate parental species do possess different mtDNA sequences). Protein electrophoresis already proved successful with butterflies, when allozymes were used as nuclear markers (Descimon & Geiger 1988; Aubert et al. 1997; Deschamps-Cottin et al. 2000). And a recent work based on DNA analysis of the nuclear gene ‘engrailed’ and the mitochondrial gene CO1 demonstrated that P. nandina is actually a hybrid between a P. dardanus male and a P. phorcas female (Thompson et al. 2011). 5 121 122 123 124 We have now combined allozyme electrophoresis with the sequencing of both nuclear and mitochondrial DNA segments in order to demonstrate that E. serotina is actually a hybrid. Moreover, we have definitely identified its parents, which, remarkably, happen to be only distantly related within the genus Erebia. 125 126 2. Material and methods 2.1. Biological sampling 127 128 129 130 Field work was conducted mainly by HD, whether in the times of the discovery of E. serotina (1953-1960), or in recent years (2000-2007). In the former period, searches were confined to the neighbourhood of Cauterets, whereas in recent years they were extended all along the Pyrenean range (tab. 1). 131 132 133 134 135 136 During the early period, the research sites were visited on every sunny day between the middle of August and the end of September. In general, the morning was devoted to the bottom of the valley, between elevations of 900 and 1050 m, and the afternoon to higher altitudes, up to 1500 - 1800 m. From the year 2000 onwards, more extensive prospecting was carried out along the Pyrenean chain, after potential sites had been located on the maps, and a number of localities between 800 and 1500 m were visited on each day. 137 138 Captured individuals were kept alive at 0°C in an isotherm bottle until deep freezing at -80°C in the laboratory. 139 140 141 2.2. DNA extraction and mitochondrial DNA procedures Individuals from which DNA was extracted are listed in Table 2. Total DNA was extracted as described in Aubert et al. (1999), following a standard phenol/chloroform procedure. 142 143 Segments of mitochondrial DNA coding for part of the large (16S) ribosomal RNA (LSU), NADH dehydrogenase subunit 1 (ND1) and cytochrome oxidase subunit 1 (CO1) were 6 144 145 146 amplified by PCR following Aubert et al. (1999), Martin et al. (2002) and Michel et al. (2008). Automatic sequencing of purified PCR products was carried out either by Genome Express S.A. or GATC Biotech AG, using the same primers as in the PCR reactions. 147 2.3 Nuclear DNA amplification and sequencing 148 149 150 151 The ITS2 segment was PCR amplified with primers ITS3 (5’- GCATCGATGAAGAACGCAGC) and ITS4 (5’-TCCTCCGCTTATTGATATGC) (Wiemers et al. 2009). PCR was carried out with Taq polymerase (New England Biolabs) in its ‘ThermoPol’ reaction buffer supplemented with 5% acetamide, 20 µM of each dNTP and 10 152 153 154 µM of each primer. Amplification was achieved with 40 cycles of 94°C/15 s, 50°C/45 s, 67°C/2.5 min and abundant material was obtained, except with the E. manto W298 extract, which repeatedly failed to yield any detectable product. 155 156 157 158 159 160 161 162 163 164 After extended electrophoresis of PCR products on an agarose (Invitrogen) gel, apparently single bands at ca 660 and ca 600 nucleotides were observed in the E. pronoe W293 and E. epiphron W300 lanes, respectively (see fig. 3A). By contrast, both E. serotina extracts (W295 and W303) reproducibly generated two bands: a major product (band B) was found to comigrate with the E. epiphron band, while a minor one (band A) migrated approximately as the E. pronoe band (fig. 3A). Each of the two bands from a W295 PCR was cut out from the gel and 1 µl of melted agarose was used for reamplification. Material from band B generated a single PCR product that migrated as the initial one. On the other hand, reamplification of band A kept yielding multiple bands: a major one that migrated as the original band A and one or several minor bands that migrated as band B or in between bands A and B (not shown). 165 166 Products from amplifications and reamplification were purified (Sigma GenElute kit) and sequencing was attempted at GATC Biotech with primers ITS3 and ITS4. Complete 7 167 168 169 170 171 172 173 174 175 176 177 178 179 sequences were readily obtained for E. serotina band B (Accession # KC184898) and E. pronoe (Accession # KC184896), despite the presence in the latter of a minor molecular population with four, rather than three GT repeats at ITS2 position 168. Reading of the E. epiphron sequence with primer ITS3 was hampered by the presence of two PCR products, with three and four TG repeats, respectively, at ITS2 position 14. However, a nearly complete sequence (Accession # KC184899) was obtained with primer ITS3epi (5’- TGTGTGTTGGACACGTGG), that partially overlaps onto the repeats. With E. serotina band A, primer ITS3 generated a sequence which could be deciphered only partially, but those segments that could be read proved identical or very similar to their E. pronoe counterparts. A clean, complete sequence (Accession # KC184897) could eventually be obtained by using a primer (ITS3pro, 5’-ATACCACACTGTCCAGCCC), the last two nucleotides of which, at ITS2 positions 31-32, matched the provisional E. serotina A sequence, but not the E. epiphron sequence. 180 2.4. Phylogenetic analyses 181 182 183 184 185 186 187 188 189 190 Sequences published prior to 1 January 2010 and claimed to pertain to Erebia and the mitochondrial genes encoding products ND1 and ND5 (NADH dehydrogenase subunits 1 and 5), CO1 and CO2 (cytochrome oxidase subunits 1 and 2) and LSU (large ribosomal RNA) were collected from databases (Appendix) and manually aligned. When several identical sequences were available for a single taxon, only one was retained. The following entries were discarded as well: (i) AY090212 (‘E. palarica Chapman 1905 isolate 9-4’, actually E. ligea L. 1758); (ii) FJ663463 (‘E. callias altajana Staudinger 1901 voucher 2005-LOWA- 581’, actually a relative of E. pharte Hübner 1804 and E. kindermanni Staudinger 1881); (iii) AY346228 (‘E. gorge Hübner 1804’, in fact almost identical to entry AY346222, belonging to E. cassioides Hohenwarth 1793); (iv) EF545697 (‘E. alcmena Grum-Grschimailo 1891 8 191 192 193 voucher H-026’, actually not an Erebia, possibly a Hyponephele); (v) AF214625 (‘E. epipsodea Butler 1868’, actually E. epiphron); (vi) AF229963 (‘E. cassioides’, actually not an Erebia). Boeberia parmenio Boeber 1809 was used as outgroup. 194 195 196 197 198 199 200 Our final alignment (available from FM), which comprises sequences from 87 individuals and encompasses a total of 4101 sites with gaps interpreted as missing data, was fed to MrBayes3.1.2 (Ronquist & Huelsenbeck 2003) as a single partition, with a GTR (General Time Reversible) model of nucleotide substitutions, gamma-distributed rate variation across sites and a proportion of invariable sites. The program was set to perform 3 000 000 iterations, of which the first 1 000 000 were discarded prior to checking convergence between the two parallel runs, calculating binary partitions and, from them, a consensus phylogram. 201 202 203 204 205 206 207 208 209 210 211 212 213 214 2.5. Allozyme electrophoresis Nine to 12 imagines from each of the putative parental species (E. epiphron, E. manto, E. pronoe) were collected in Vallée du Lys (locality D), while the imago of E. serotina was from the neighbouring Granges d’Astau (locality E); they were stored at –80°C until electrophoresis. Five loci were examined for each sample: Glutamate-oxalacetate dehydrogenase (GOT, EC 2.6.1.1), Isocitrate dehydrogenase (IDH, EC 1.1.1.42), Phosphoglucose isomerase (PGI, EC 5.3.1.9), Phosphoglucomutase (PGM, EC 2.7.5.1) and Mannose-6-phosphate isomerase (MPI, EC 5.3.1.8). Electrophoreses were carried out on cellulose acetate gels according to Hebert and Beaton (1993), with slight modifications. TrisGlycine buffer (25 mM Tris, 192 mM Glycine, pH 8.5) was used for MPI, PGM, PGI and GOT, and Tris-citrate buffer (50 mM Tris, 12.5 mM citric acid, pH 7.8) for IDH. Gels were run at 200 V during 25 min for MPI, 40 min for PGM, 50 min for GOT, 60 min for IDH and 75 min for PGI. 9 215 216 Genotypes of all the individuals were determined. Several individuals were run at least twice to check the repeatability and the relative position of all alleles in the different species. 217 218 219 220 3. Results and Discussion. 3.1. The pursuit of E. serotina: a narrative 3.1.1. Early observations (1953-1960) 221 222 223 224 225 226 227 Around Cauterets, most of the explored areas lie in the ‘montane’ zone, where beech and fir forest is the climax vegetation. All E. serotina individuals were collected at only two sites, one ca 0.5 km upstream of Cauterets (Pont de la Raillère, locality A) and one ca 1.5 km downstream (Catarrabes and Canceru hamlets, locality B). Both places consisted of flowery meadows; the former was lightly grazed and the second was mowed once or twice a year. At the time of the visits, sites at higher altitudes occasionally yielded old individuals of other Erebia species, but no E. serotina. 228 229 230 231 232 The two sites look rather different. The Pont de la Raillère is deeply set between two very steep mountain slopes, while the Catarrabes-Canceru meadows lie within a larger, sunnier valley floor. However, these places have in common that they look somewhat like the bottom of a funnel, the lower slopes of which are covered with forest and the upper ones with subalpine grassland. 233 234 3.1.2. Recent observations (1999-2007) 235 236 Pont de la Raillère and Catarrabes (Localities A and B) were visited in 1999 and 2002, with no success; the site itself underwent little change, but some of the neighbouring meadows had 10 237 238 become afforested. The Catarrabes-Canceru hamlets had become more urbanized and the meadows were out of reach. 239 240 241 242 243 Site C had been discovered in the Arrens valley, about 20 km west of Cauterets, by LouisAugustin (1985 and pers. comm. to HD), who collected there two E. serotina individuals in the 1970s. That site was visited by HD in September 1999, no E. serotina was captured, but the characteristics of the locality were found to be similar to those of the Cauterets neighbourhood: flowery meadow around 1000-1200 m elevation, funnel-bottom-like site. 244 245 246 247 248 249 250 251 252 253 254 Following these setbacks, more effort was devoted to explore the eastern part of central Pyrénées, namely the Aure and Luchon valleys. A favourable locality (D) was spotted on the floor of the Lys valley, south of Luchon, and five E. serotina individuals were collected there between the 5th and 8th of September 2000 (two of these individuals were further selected for molecular analysis under the references W295 and W303). The funnel-like overall topography of this site was obvious and extensive; moderately grazed and flowery meadows proved attractive to butterflies. The locality was visited again in early August 2001, in order to collect individuals of the putative parent species. On this occasion, it was observed that only E. meolans and E. euryale were to be found at the very place at which E. serotina had been captured in 2000. On the other hand, all three putative parental species were abundant above the continuous and extensive forest belt, in the lower subalpine grasslands. 255 256 257 258 259 260 In 2002, Vallée du Lys (site D) was explored again in the first half of September, with no success. However, an individual of E. serotina (further checked for allozymes under the reference EM) was found westwards of that site, in the neighbouring valley of Oo, once again in a typically funnel-shaped site, at the Granges d’Astau meadows (site E). The same year, M. J. Lalanne Cassou (pers. comm.) found an individual in the lower part of the Lys valley, close to Bagnères de Luchon, in early October.

This Is an Author's Original Manuscript of an Article Whose Final And

Details

Download

Copyright

We respect the copyrights and intellectual property rights of all users. All uploaded documents are either original works of the uploader or authorized works of the rightful owners.

Support